Who Owns the Future? (38 page)

When Will Decisions Be Made?

It would be humanly impossible for a person to constantly make all the decisions needed in an advanced information economy. Say you want a cab. Today it is already possible to call a cab with a smartphone, but to do so a user has to have populated a stack of about ten interdependent decisions.
*

*
A phone must be chosen, and a wireless carrier, and a payment service, and a taxi-calling app, and an email account to tether the payment and taxi apps, and a credit card to process the payments, and a bank to tether the credit card to, and possibly a PC to tether the phone to, and connectivity for home Wi-Fi, and a contacts management app or social network to keep track of the addresses of places the taxi might take you.

Ten is a lot. Such a tangle of decisions can only be reconsidered infrequently. In some cases the decisions are forced,
^†
which is annoying, but also a cognitive benefit in disguise.

†
Choosing a phone might force the choice of wireless carrier, for instance.

It isn’t hard to imagine future scenarios in which the stack could grow to hundreds or thousands of decisions. That would certainly be possible when you bring an elder-care robot into your life, or operate your 3D printer.

Any desirable alternative economic future must include an idea about a user interface that brings at least as much simplicity to people as acquiescing to a Siren Server does today. This means reducing the density of decisions people are expected to make to a level that leaves cognitive room to live life in free and creative ways.

If Siren Servers turn out to be the only way to reduce the burden of
decision making in an information economy, then we are done. That would mean there is only one possible design for high-tech society.

However, there are almost certainly other options. Imagine a future industry of “decision reduction” that would be (gasp!) regulated so as to remain unaligned with other services. You’d choose a decision reduction service the way you choose a broker now. The decision reduction service would use its particular style and competence to create bundles of decisions you could accept or reject en masse. You could switch to other services without penalty at any time. Such services would be prohibited from having conflicts of interest. That is a proper place for regulation.

A little basic regulation would force decision reduction services to be competitive instead of being vulnerable to the moral hazard of locking people into contracts. This idea is a generalization of many familiar ideas from antitrust and network neutrality.

If we allow ourselves to lean into a utopian stance just a little, then we can suppose that the ideal solution would be an open market in decision reduction, which even individual entrepreneurs could enter. Just like a personal assistant, a certain sort of person might be effective and happy reducing the choice space for others. In other cases, delegation to a huge decision reduction cloud service worth hundreds of billions of dollars might be the best choice for a particular customer.

The possibility of new kinds of personal assistants adds to the arsenal of answers to the question “What would people do?” In a world of thorough and honest accounting, whole new large classes of service professions should naturally pop up.

In early experiments like Second Life, we’ve already seen glimmers of new paid roles, from avatar stylists to virtual performance venue promoters. Facebook and the like also generate fledging new paid roles, but they’re often defensive and dreary, like reputation protection and restoration.

Once a humanistic economy gets going, I imagine that accounting will suddenly become an interesting job. Accountants will be called upon to expand the kinds of value that can be documented to enhance the network. They’ll not only get their clients paid, but also cause the economy to grow. They’ll be a little like politicians and a little like detectives. They will not be backroom nerds but action heroes.

New careers as fresh as these, or beyond my imagination, should be appearing already, but the Sirenic pattern shuts down that kind of progress.

If I try to imagine what it would be like to be an individual in a humanistic economy, I suppose that a big life choice would be how much attention to devote to one’s information transactions. One choice would be to be lazy in the quotidian sphere of life and sign up with a decision reduction service, but then double down on whatever you’re good at that generates your income and wealth. Another choice would be to become personally obsessed with the details of your information life. People with a mind to do it could optimize their information incomes and wealth creation, but might as a result not look at the big picture as much. There would be all sorts of in-between options to suit different personalities.

Once again, as a reminder, this argument is neither anticorporate nor redistributionist. The test of success ought to be that both the big players and individuals do better in a growing economy. To put it another way, there ought to be big corporations doing big jobs without necessarily having to become Siren Servers.

Dynamic Value

The price of computation in a humanistic information economy ought never be set exclusively by rote, but always be determined to a significant degree by market negotiation. We will never know for sure in advance how valuable a particular datum might turn out to be. Each use of data will determine a fresh valuation of it in context.

There will be vastly more commercial events than in the world we are used to. Every time code runs, a lot of people will be paid a tiny bit each. There is no such thing as calculation without data. Therefore, if the provenance of the data has been preserved, then calculations can generally be expanded to yield additional results about who should get credit for making them possible.

It will be very rare, essentially impossible, for Amazon to sell a book for zero dollars, as it sometimes does today. This is because it will be almost impossible to assemble any information stream for
which no component has some established value, or for which there is no potential customer. In physicality, it isn’t unusual to see puppies or large items offered for free, because it’s hard for the owner to keep them. That’s almost never the case for information. There should be far less free stuff in an information economy than in physicality.

There will be no upper bound to a price. Sellers at every level will be able to set prices as high as their markets will bear, but competition will keep prices in check.

The principle would apply to code as well as data. Computer code these days tends to be either proprietary or open-source. A third option would come into being in the future proposed here, and perhaps into ubiquity. Code would remember the people who coded each line, and those people would be sent nanopayments as part of code execution. A programmer who writes code everyone uses will be able to benefit directly, instead of having to leverage code into a Siren Server scheme. The Google guys would have gotten rich from the search code without having to create the private spying agency. At the same time, an open community of programmers would have been able to contribute incrementally, without any more barriers than are found in today’s open-source community.

My current thinking, which will undoubtedly not be the last word, is to calculate prices in humanistic transactions in a mixed way, partially determined by buyers and sellers in the moment and partially determined automatically by universal policies. Each price will have two components, called “instant” and “legacy.”

The reason for this is to account for the value that people have already brought into the world. Capitalism has suffered from a memory disorder. It’s been so glued to the moment, to the current deal on the table, that it’s possible for an economic crash to occur in the midst of wealth.

The “instant” part of the price will arise from agreement between buyer and seller. Just as in physicality, there are varied mechanisms by which agreement can be reached. Sometimes a seller will set a retail, take-it-or-leave-it price. Or there might be an auction, or a back-and-forth negotiation. A mature information economy ought to spawn new styles of price determination. We’ll talk more about this interesting new frontier shortly.

The “legacy” portion of the price will be composed of algorithmic adjustments to instant pricing that uphold the social contract and economic symmetry. Here are examples of the sorts of legacy adjustments that might be incorporated:

• Something old: Tax.

• Something new that might make a price go up: Calculation of the relative contributions of upstream people to the value of the transaction, so that they will be compensated. A buyer and seller can’t set a price to screw over those who came before, without whom the present transaction would be impossible. Those who came before remain first-class economic citizens, though they must contend with market forces and can’t set arbitrarily high prices. The next adjustment will prevent them from engaging in “blackmail” pricing.

• Something new that might make a price go down: Incremental correction for examples of software lock-in or other impediments to competition, so that antitrust-like problems are avoided in advance. This would not be a matter of bureaucratic judgment, but a dispassionate mathematical calculation. The calculation would answer the question “How much would it cost the buyer if prior decisions about ‘populating the stack’ had been different?”
*
For instance, suppose you had chosen a different wireless service in the past and want to call a cab with your smartphone now. If that different past decision in how you populated your stack would have caused a
major
difference in the cost of calling the cab now, larger than some threshold, then that should be understood as an instance of unproductive lock-in. The price paid would partially, not totally, be adjusted to undo the moral hazard of lock-in. In the present system, businesses need to rely on lock-in to make a profit in the online world, but in the world foreseen here there would no longer be a proper function for it.

*
The notion of calculating “What if things had been different?” ought to alarm mathematically inclined readers. Would it always be possible to calculate counterfactual financial histories? Wouldn’t there be many chaotic situations in which petty differences would have had huge implications? This would indeed become a major area of concern in a humanistic information economy. It is beyond the scope of this book to go into the topic in detail, but the key idea is to design an economy to incentivize and otherwise foster more “linear” financial dealings that avoid chaos as much as possible. When the answer to “What if things had been different?” is chaotic and mostly meaningless, then chances are that the actual happening was also thus. The point is to make capitalism as little like a casino as possible.

The legacy portion of transactions overall might be centrally regulated. If it’s too low, it won’t act as a flywheel, propelling the economy forward to make sure it doesn’t stall or fall into Keynesian traps. If it’s too high, it will undo the motivational aspect of the market, since outcomes would then be based too much on what happened long ago.

Earning a Little Money by Living Well or Interestingly

Here’s a simple example of how you might make money from the cloud in a humanistic future of more complete accounting. It’s based on the kind of dubious calculation that’s typical of cloud entrepreneurship today.

You meet a future spouse on an online dating service. The algorithms that implement that service take note of your marriage. As the years go by, and you’re still together,
*
the algorithms increasingly apply what seemed to be the correlations between you and your spouse to matching other prospective couples. When some of them also get married, it is automatically calculated that the correlations from your case were particularly relevant to the recommendations. You get extra nanopayments as a result.
^†

*
Wishful thinking, I know. Investors in today’s startups want their money out so soon that the services they fund can’t even be tested against the rhythms of human life. Hopefully that will change.

†
This process would be a microscopic echo of what is already done when you get a mortgage. By paying it you not only create new money, but you strengthen property values around your own house, effectively making a little of the money that your neighbors create when they get mortgages.

This sort of result is already calculated today, but the payments don’t flow. The extra work for the microprocessors in the cloud computers would be trivial, considering the expected course of Moore’s Law, and the extra payments would expand the economy
for everyone, including the cloud computing companies. Economic expansion ought to more than pay for the extra trouble.

Would the correlation be valid? Well, this would be business and not science. Honestly, as I explained earlier, I am super-skeptical of algorithms of this kind. It’s incredibly hard to design experiments that separate the influence of such algorithms from their predictive veracity. They create their own validity if people are willing to use them. That critique is economically irrelevant, however. The point is that if future couples pay for a service for which you and your spouse contribute data, you’ll benefit—though only proportionally.

This brings us to the “instant” part of the calculation of the nanopayment to you. It should be proportional to
both
the importance of the data that came from your state or behavior
and
what the seller downstream was able to earn
and
whatever profit you or your decision reduction partner tried to extract. So, for instance, if the dating service were due an extra fee for brokering a successful marriage, part of that fee would go to you.